Time-Varying Dictionary and the Predictive Power of FED Minutes

Lima, Luiz Renato; Godeiro, Lucas Lúcio; Mohsin, Mohammed

doi:10.1007/s10614-020-10039-9

Cited by 8 publications

(10 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(iii) Our approach considers fixed content which can be made dynamic (as in Lima et al (2020)) and compatible with forecasting, (iv) Variances of financial variables are generally time-varying. Splitting the period helps limiting these effects, but modelers often use GARCH/ARCH, i.e.…”

Section: Discussionmentioning

confidence: 99%

Tracking ECB’s Communication: Perspectives and Implications for Financial Markets

Fortes

Guenedal

2021

SSRN Journal

View full text Add to dashboard Cite

This article assesses the communication of the European Central Bank (ECB) using Natural Language Processing (NLP) techniques. We show the evolution of discourse over time and capture the main themes of interest for the central bank that go beyond its traditional mandate of maintaining price stability, enlightening main concerns and themes of discussion among board members. We also built sentiment signals compatible with any form of language, both formal and informal, an important step as the ECB aims to enhance communication with non-expert audiences. In a second step, we measure the impact of the ECB's communication on the EUR/USD exchange rate. We found that our quantitative series, both topics and sentiment, improve financial-linked models consistently in all periods analyzed (2.5% on average). Meaningful signals comprise a broad range of subjects and vary in time. This suggests that overall ECB's talk matters for asset prices, including themes not directly related to monetary policy. This result is particularly important in a context in which the ECB, as well as other major central banks, are moving towards integrating issues closer to the society into their scope of action, implying that subjects, which were considered peripheral, may become central. This emphasizes the importance for markets to effectively track central banks' communication to improve investment processes.

show abstract

Section: Discussionmentioning

confidence: 99%

Tracking ECB’s Communication: Perspectives and Implications for Financial Markets

Fortes

Guenedal

2021

SSRN Journal

View full text Add to dashboard Cite

show abstract

“…The optimal values of

{\theta}_1

and

{\theta}_2

are obtained from the procedure suggested in the GLMNET R package (linear regression section) developed by Trevor Hastie and Junyang Qian. In the forecasting literature, a short list of papers that successfully employed elastic net includes Bai and Ng (2008), Li et al (2015), Lima et al (2019), among others, whereas Lima et al (2019) utilizes elastic net to select the most predictive words from minutes published by the federal open market committee (FOMC).…”

Section: Methodsmentioning

confidence: 99%

Equity‐premium prediction: Attention is all you need

Lima

Godeiro

2022

J of Applied Econometrics

View full text Add to dashboard Cite

Summary Predictions of stock returns are greatly improved relative to low‐dimensional forecasting regressions when the forecasts are based on the estimated factor of large data sets, also known as the diffusion index (DI) model. However, when applied to text data, DI models do not perform well. This paper shows that by simply using text data in a DI model does not improve equity‐premium forecasts over the naive historical‐average model, but substantial gains are obtained when one selects the most predictive words before computing the factors and allows the dictionary to be updated over time.

show abstract

“…To meet the time‐varying characteristics of texts (Lima et al, 2020), we conduct a recursive window to estimate and update our text‐based PBC communication indices, which makes it more efficient to use all data when the parameters do not change substantially over time. It is also widely used in a number of forecasting studies (Andreou et al, 2013; Ellingsen et al, 2021; Stock & Watson, 2006).…”

Section: Text‐as‐datamentioning

confidence: 99%

“…We use the data from 2003Q1–2009Q4 (or 2003:M1–2009:M12) as the initial window to estimate the predictors and train the forecasting models, and we recursively update and expand them for the OOS evaluation. Following the general settings (Lima et al, 2020; Lin et al, 2021), we consider both short‐ and long‐term forecasting horizons. More concretely, we set

τ = 1, 2, 3, 4

for quarterly targets (e.g., GDP growth rate) and

τ = 1, 3, 6, 12

for monthly targets.…”

Section: Forecasting Modelsmentioning

confidence: 99%

“…Our second contribution is emphasizing the importance of real‐time textual analysis in economic applications to avoid look‐ahead biases, especially in macroeconomic projections. For our real‐time forecasting application, a key issue is to avoid look‐ahead biases (Larsen et al, 2021; Lima et al, 2020; Thorsrud, 2020), which is also important for textual analysis but often ignored in the literature (Kelly et al, 2021; Lin et al, 2021). Our work refines HDMR to construct the PBC communication index in real time and documents that the OOS performances are to some extent different (especially for nowcasting), even though the full‐sample and real‐time estimates for text‐based indices are analogous.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Real‐time macroeconomic projection using narrative central bank communication

Lin

Fan

Zhang

et al. 2022

J of Applied Econometrics

View full text Add to dashboard Cite

Summary Unlike the central banks of most developed economies, the People's Bank of China (PBC) does not release its macroeconomic forecasts to the public but instead carries out narrative communication. We apply a hurdle distributed multinomial regression to PBC communication texts in real time, addressing the ultrahigh dimensionality, sparsity, and look‐ahead biases. In addition, we embed text‐based indices into mixed‐data sampling (MIDAS)‐type models and conduct forecast combinations for prediction. Our results argue that the predictive information from communication texts improves the real‐time out‐of‐sample prediction performance. We connect textual analysis and real‐time macroeconomic projection, providing new insights into the value of central bank communication.

show abstract

Time-Varying Dictionary and the Predictive Power of FED Minutes

Cited by 8 publications

References 40 publications

Tracking ECB’s Communication: Perspectives and Implications for Financial Markets

Tracking ECB’s Communication: Perspectives and Implications for Financial Markets

Equity‐premium prediction: Attention is all you need

Real‐time macroeconomic projection using narrative central bank communication

Contact Info

Product

Resources

About